It’s done! After four months of intensive construction, 3D printing, assembly and setup, my second 3D-printed and self-built mini observatory is finally finished. With a diameter of 80 cm, it is almost twice as high as my first mini observatory built in 2022, with which I took so many beautiful astrophotos. I enjoyed the convenience of not having to demount a scope at the end of the session, in the night or early in the morning. Now I can take photos even on nights that are not completely clear. In my first mini observatory, I could only mount a lens with a maximum focal length of around 110 mm. Also, operation with a ZWO ASIAR Pro was not really good for longer unattended imaging sessions. Above all, the ASIAIR software lacks useful functions for adjustments or corrections in the event of unexpected disturbances, such as star loss in guiding when clouds pass through. And the autofocus was unfortunately not always precise enough.
I no longer wanted to modify or extend the mini observatory. I would rather build a new, larger observatory. After evaluating various construction options, my experience again led me to decide that the best solution was a clamshell observatory, an observatory with a dome made of two movable half-shells. And it was also clear that the observatory should now be operated under Windows with the versatile open source software N.I.N.A..
The dome of the observatory simply had to be as large as possible. In the end, however, it is determined by a reasonable 3D-printing effort, i.e. primarily by the size of the print bed and the printing speed of the 3D printer. Just for fun I had estimated the pure printing time for my first 3D printer, a Prusa MINI to be around 150 days. Certainly unthinkable, far too long! But the K1 Max 3D printer from Creality would only need around 15 days of pure printing time with a Hyper Filament and its print area of 30 cm x 30 cm..
Building the Dome
Thought, done. With my experience, building this second observatory was no longer a dramatic challenge, but rather an exciting routine. With basic knowledge of Autodesk Fusion365, which is free to use for private users, the full-size dome was quickly designed. There were essentially six full-size parts: a base ring, an intermediate ring and four movable half-shells for opening and closing the dome. There were also various smaller parts, brackets for the motors and pulleys, axles, cable channels and covers.
As I couldn’t print parts with a diameter of just over 80 cm in one piece, I used a slicer program to cut the base and intermediate ring into eight parts, and each half-shell into nine. For stability, I designed and printed two half shells for each side. This allowed me to slice them at different positions so that I could place the two elements on top of each other and screw them together at the seams.
Four weeks after printing began, all the substantial dome parts had been printed. Twelve kilograms of Hyper-PETG plastic had been used. Parts were then put together, glued or screwed together. During printing, however, some parts of the dome shells had warped slightly, leaving a handful of small gaps of around 1 to 2 mm when put together. They were quickly closed with 2k filler. And finally, the entire dome was painted white with acrylic paint as better protection against UV radiation. A strip of aluminum sheet on the inside of the base ring provides additional stability. However, the base ring is not evenly round. It is slightly curved inwards on the south side. This means that the corresponding dome shell opens slightly wider than the other, and the horizon for the telescope is about 10° lower to the south. Whether it was necessary for me, I don’t know. Ultimately, however, the surrounding trees and bushes determine the height of my usable horizon for the telescope. With a few exceptions, it is around 30°.
The observatory should actually be watertight. The half shells overlap by about 5 cm and the screws in the shells are sealed as well as possible with silicone. Nevertheless, during longer periods of bad weather, I cover the observatory with a specially designed round cover. And for my peace of mind, there is an alarm system. It secures the observatory against unauthorized access or theft.
Motorization
Small, 12V DC motors should move the shells and open and close the dome. Initially, I opted for the visually very appealing winch motors for model cars on a scale of 1:10. These motors can certainly lift each of the four-kilogram half-shells. Unfortunately, however, they are not self-locking, i.e. the weight of the half-shells causes the switched-off motors to rotate backwards and the dome to open automatically. So I chose small motors with worm gears instead, which are self-locking by design. A 0,5 mm thin wire rope is rolled onto the drum on the motor and runs over the outside of the intermediate and base ring to the lower end of the half-shell via a pulley. There it is bolted to the half-shell so that the motor pulls it up to close the dome. It groans a little bit, but the motors do their job quite well..
An Arduino Uno with an L298N module controls the two motors via button or ASCOM interface. This allows them to rotate independently of each other and in both directions. For ASCOM control from N.I.N.A. I use the RRCI-ASCOM dome driver with an Arduino sketch adapted for my clamshell design. I only got the RRCI driver to work on my ATmega Uno, it didn’t work with my other Arduino processors. Two microswitches report the end positions of each half-shell to the microprocessor to stop the motor. Since I didn’t want to rely solely on the limit switches, the motors are also stopped automatically two seconds after their individual run time.
Inside the Observatory
In this dome I finally had space for my AM5 mount and the Askar SQA55 telescope with 264 mm focal length, my ASI183MC Pro camera with rotator and filter wheel, a focuser and an ASI120MM camera with a 130 mm / f4 guiding telescope. These parts weigh six kilograms together. But thanks to the AM5 strainwave mount, I can forego any counterweight that is necessary for classic mounts. And the dome has another very welcome feature: the dome can be closed in any position of the telescope.
The necessary electronics and cabling are arranged in such a way that the complete mount with telescope and attached accessories can be inserted into or removed from the dome in about 5 minutes. For this reason, a mini PC, USB hub and an Arduino microprocessor for controlling two anti-dew heating bands are mounted directly to the telescope or its mounting rail. They rotate with the telescope in any direction and thus avoid a tangle of connecting cables when moving the telescope.
Controlling the Observatory
The observatory can be controlled completely remotely. It is is integrated into my local network and can also be accessed from the Internet via VPN.
I have attached a network-controlled Waveshare 8-channel relay to the mount, which can switch on and off the power supply of the individual components, including a white LED to illuminate the dome in a case of emergency. The relay can be operated via an ASCOM driver, but also via the Home Assistant app, a home automation control software. The dome itself also houses the ATmega microprocessor mentioned above for controlling the dome motors.
Self-regulated dew heater
To prevent dew from fogging the lenses of the telescope and guiding telescope, I have built a self-regulating dew heater. It is based on the Seeeduino-WiFi microprocessor, which is only the size of a thumbnail. Depending on the dew point and the temperature change at the telescope it regulates the temperature of two heating bands.
The data required for this is provided by a DHT22 sensor and a temperature-dependent PTC resistor. The DHT22 measures temperature and humidity, while the PTC resistor measures the changes in temperature on the telescope. The microprocessor uses this data to regulate the power of both heating bands via PWM modules. For control purposes, it also measures the current of the heating bands with an ACS712 module and provides this measured data on a webpage.
There are also some other components that make operation easier. A small wide-angle camera is located inside the dome to monitor the dome, telescope and mount. A network switch connects the PC, the surveillance camera and the 8-channel relay to my network.
Final Words
I can’t deny it, it was a lot of work to build, test and optimize the observatory. But I also had a lot of fun over the winter. The astrophotography suite N.I.N.A. is so great. The autonomous recording sessions run smoothly. All the effort was worth it.
Finally, a few key data. The field of view of the telescope system is about 2° x 3°, corresponding to 2″/pixel. The guiding is a good 0.5“ to 0.8”, which is completely sufficient for this setup. The Askar SQA55 telescope has excellent optics. The images are extremely sharp, the stars are faultless right into the corners.
Now in spring it is still galaxy time, the Milky Way with its colorful nebulae only rises again shortly before dawn in the northern hemisphere. And my observatory with its field of view of around 2° x 3° is predestined for these objects. I wish myself many clear nights for this.
Related Links
ASCOM RRCI Dome Driver
https://projecthub.arduino.cc/cfar/rolling-roof-computer-interface-rrci-a7f9ac
ASCOM Driver to control the Waveshare relay
https://github.com/ngaertner/ASCOM-Switch-Waveshare-POE-ETH-Relay